Spin- and charge density around Rh impurity in α -Fe studied by Mössbauer spectroscopy

Transcription

1 Spin- and charge density around Rh impurity in α -Fe studied by Mössbauer spectroscopy A. Błachowski 1, K. Ruebenbauer 1* and J. Żukrowski 57 Fe 1 Mössbauer Spectroscopy Division, Institute of Physics, Pedagogical University PL-3-84 Cracow, ul. Podchorążych, Poland Solid State Department, Faculty of Physics and Applied Computer Science, AGH University of Science and Technology PL-3-59 Cracow, Al. Mickiewicza 3, Poland * Corresponding author: sfrueben@cyf-kr.edu.pl PACS: 75.5.Bb, 75.3.Fv, Lr, y Keywords: spin and charge density, Mössbauer spectroscopy, impurities in metals Short title: Spin- and charge density in α -FeRh Accepted on October 7 th 8 in The Journal of Alloys and Compounds as a regular paper [Section: Physics] Submitted on September nd 8 Journal of Alloys and Compounds 477(1-), 4-7 (9) May 7 th 9 DOI: Abstract Random solution of rhodium in ferromagnetic α -Fe of the BCC structure has been 57 investigated by means of the 14.4 kev Mössbauer transition in Fe at room temperature. Rhodium atoms have been randomly substituted on the iron sites with the concentration up to 15 at.%. Contributions to the hyperfine field and isomer shift on the iron nuclei have been determined as the function of the distance between iron nucleus and rhodium impurity up to the third co-ordination shell. Rhodium atom as the nearest iron neighbor changes iron hyperfine field by +.73 T, as the second neighbor makes change by +.7 T and finally as the third neighbor changes the field by +.43 T. Corresponding changes in the isomer shift are as follows:.15 mm/s,.5 mm/s and +.39 mm/s. The average hyperfine field and the isomer shift increase linearly versus rhodium concentration at rates. 156 T/at.% and x 1 mm/(s at.%), respectively. Hence, addition of rhodium lowers the electron density -3 1 ρ on the iron nucleus at the rate ρ / c = 1. 5 x 1 electron a.u. (at.%). 1

2 1. Introduction Mössbauer spectroscopy is sensitive to the local environment of the resonant atom. In particular one can investigate influence of the adjacent impurity on the hyperfine field and isomer shift experienced by the resonant nucleus [1, ]. Perturbations to the spin- and charge density on the iron nucleus due to the impurity located on the regular lattice site of the BCC iron could be seen to the second or sometimes to the third co-ordination shell [3, 4]. Rhodium dissolves randomly in the BCC α -Fe up to about 11 at.% at low temperature equilibrium [5]. However, the α -phase is stable at low temperature up to about 55 at.% of rhodium. For the rhodium concentration range at.% one can observe a tendency for ordering into the B structure. No ordering occurs for rapidly quenched samples with the rhodium concentration less than 19 at.% [5]. Chemically disordered α -phase is ordered ferromagnetically at room temperature. The Curie temperature decreases with the rhodium addition, but it is still quite above the room temperature even for the most rhodium abundant disordered BCC phase. The iron-rhodium system has been investigated by the Mössbauer spectroscopy in the past [6-9], but usually for the high rhodium concentrations or for the rhodium concentrations leading to the ordered B phase. We have investigated random solutions of rhodium in the α -Fe for the rhodium concentration up to 15 at.% by means of the room temperature Mössbauer spectroscopy.. Experimental Samples were prepared by arc melting of the appropriate amounts of the constituent elements under protective argon atmosphere. Natural iron of the at.% purity and rhodium of at.% purity were used to make samples. Samples of about 1. 5 g were prepared by melting constituents three times to assure ingot homogeneity. Rapid cooling of the samples from the melt assured random distribution of the rhodium impurity. No weight losses were observed during sample preparation, and therefore the starting composition could be treated as the resulting sample composition. Mössbauer absorbers were prepared as powders embedded 57 in the epoxy disks with about 3 mg Fe/cm. A commercial Co(Rh) source was used. Transmission geometry with the source and absorbers kept at room temperature has been applied. Raw spectra were collected in 496 channels in the round-corner triangular mode with the help of the MsAa-3 spectrometer [1, 11]. All spectral shifts are reported versus room temperature total shift in α -Fe. 3. Data evaluation Folded spectra were evaluated within the model described previously [4, 1]. A transmission integral approximation was used. The model depends on the following assumptions. It is assumed that Rh impurities are located at random on the regular iron sites within the BCC structure. Furthermore, it is assumed that perturbations due to various impurities are additive in the algebraic sense. Those perturbations depend only on the distance between impurity and the resonant iron nucleus. Hence, one can define a contribution to the iron hyperfine field B s caused by the impurity in the s - th co-ordination shell of the resonant atom, and a corresponding contribution Ss to the isomer shift. The perturbation of the hyperfine field is due to the perturbation of the spin density on the iron nucleus, while the perturbation of the isomer shift is caused by the corresponding perturbation of the electron charge density. Subsequent co-ordination shells are taken into account till the most distant discernible shell, i.e., s =1,,...,. Shells beyond the last shell taken into account contribute to the

3 remaining hyperfine field B and spectral shift S. The average field < B > and average shift < S > could be obtained in straightforward way within this model. Essential results for = and = 3 models are shown in Table 1. Figure 1 shows spectra fitted within the = 3 model. Rhodium doped samples contain traces of the iron oxide, and a contribution of this oxide to the spectra has been taken into account during data processing. The oxide subspectrum is described by a doublet, the latter having total shift +.3(1) mm/s versus room temperature α -Fe, splitting. 34(1) mm/s, and a contribution to the total absorption crosssection of 3. 5() % averaged over various samples. On the other hand, Figure shows evolution of the model parameters versus Rh concentration c. Data were independently fitted to the Hesse-Rübartsch model [13, 14] in the thin absorber approximation and for the distribution of the hyperfine field B. Table summarizes the average field < B > and the average shift < S > obtained by this method versus Rh concentration. Distributions of the hyperfine field obtained by means of the = model, = 3 model, and by the Hesse- Rübartsch method are shown in Figure 3 for selected Rh concentrations. Fig. 1 Mössbauer transmission spectra are shown for various rhodium concentrations. 3

4 Table 1 Essential results obtained within models with = and = 3. The last row for each model shows respective averages, where appropriate. The averages have been calculated for the rhodium concentration c ranging from 1 at.%. Symbols < B > and < S > stand for the average field and shift, respectively. Symbols B and S denote contributions to the field and shift due to the atoms beyond the th co-ordination shell. Symbols B 1, B and B3 stand for contributions to the field caused by the rhodium atom located as the first, second and third neighbor, respectively. Symbols S 1, S and S 3 denote corresponding contributions to the shift. c < B > B B 1 B B 3 < S > S S 1 S S 3 (at.%) ±.1 ±. ±. ±.3 ±. ±.3 ±. ±. ±.3 ±. ± = = Table Average hyperfine field and spectral shift obtained by means of the Hesse-Rübartsch method versus rhodium concentration. c (at.%) ±.1 < B > ±. < S > ±

5 Fig. Parameters of = and = 3 models plotted versus rhodium concentration. Straight lines are obtained including data for the rhodium concentration up to 1 at.%. The symbol ρ stands for the average electron density on the iron nucleus, while the symbol ρ Fe denotes corresponding electron density in the pure α -Fe. Fig. 3 Hyperfine field B distribution within: =, = 3 and Hesse-Rübartsch models plotted for selected rhodium concentrations. 5

6 4. Discussion of results It is obvious that the = 3 model gives better results than the = model as absolute values B of the derivatives expressed as c and S c are much smaller for = 3 than for = (see Figure ). The average hyperfine field increases linearly versus increasing rhodium concentration with the slope. 156 T/at.%. The isomer shift behaves similarly with the slope x 1 mm/(s at.%). Hence, one can conclude that addition of rhodium lowers the -3 1 electron density ρ on the iron nucleus at the rate ρ / c = 1. 5 x 1 electron a.u. (at.%) [15]. The average electron spin density on the iron nucleus is increased by addition of rhodium. Therefore, the average spin density transferred from the neighborhood of the iron atom is decreasing with increasing rhodium concentration. The electron and electron spin density changes remain proportional one to another within the concentration range of rhodium investigated here, i.e., up to about 1 at.% of rhodium. Vincze and Campbell [1] investigated two samples with the rhodium concentrations 3 at.% and 5 at.%, respectively. They found that the rhodium atom increases the iron hyperfine field by T as the nearest neighbor of the iron atom. On the other hand, they found that the rhodium atom increases this field by. 79 T as the second neighbor. The latter value is in good agreement with our results. The behavior of the isomer shift and hyperfine field versus rhodium concentration is similar to the behavior found by Shirane et al. [6]. References 1. I. Vincze and I.A. Campbell, J. Phys. F: Metal Phys. 3, 647 (1973).. J. Cieślak and S.M. Dubiel, J. Alloys Compd. 35, 17 (3). 3. J. Korecki and U. Gradmann, Phys. Rev. Lett. 55, 491 (1985). 4. A. Błachowski, K. Ruebenbauer, and J. Żukrowski, Phys. Scripta 7, 368 (4). 5. T.B. Massalski, Binary Alloy Phase Diagrams, nd edn. (ASM International, Materials Park, Ohio, USA, 199), Vol. 1, p G. Shirane, C.W. Chen, and P.A. Flinn, Phys. Rev. 131, 183 (1963). 7. B. Window, G. Longworth, and C.E. Johnson, J. Phys. C: Solid State Phys. 3, 156 (197). 8. G. Filoti, V. Kuncsea, E. Navarro, A. Hernando, and M. Rosenberg, J. Alloys Compd. 78, 6 (1998). 9. A.R. Yavari, E. Navarro, H. Mori, H. Yasuda, A. Hernando, and W.J. Botta, Philosophical Magazine 8, 1779 (). 1. R. Górnicki, A. Błachowski, and K. Ruebenbauer, Nukleonika 5, S7 (7). 11. R. Górnicki RENON [ ] 1. A. Błachowski, K. Ruebenbauer, and J. Żukrowski, phys. stat. sol. (b) 4, 31 (5). 13. J. Hesse and A. Rübartsch, J. Phys. E: Sci. Instrum. 7, 56 (1974). 14. G. Le Caër and J.M. Dubois, J. Phys. E: Sci. Instrum. 1, 183 (1979). 15. U.D. Wdowik and K. Ruebenbauer, Phys. Rev. B 76, (7). 6